We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
拉曼光谱结合机器学习识别肥皂.
- Authors
侯赛文; 李春宇; 孔维刚; 刘金坤; 屈音璇
- Abstract
Raman spectroscopy and statistics were used to identify different brands and types of soap accurately and quickly. The Raman spectra of 56 kinds of soaps from different manufacturers and brands were collected, and the samples were divided into five categories through Raman spectrum preprocessing and cluster analysis. In the use of SVM, KNN, Bayesian method to establish discriminant analysis, the accuracy of the test discrimination comparison, comprehensive comparison of KNN model classification effect is the best, the classification accuracy is 96.4%.This method can be used to test soap samples quickly and accurately, which provides a new method for the identification of soap samples in the process of handling cases.
- Publication
Applied Chemical Industry, 2022, Vol 51, Issue 1, p281
- ISSN
1671-3206
- Publication type
Article
- DOI
10.16581/j.cnki.issn1671-3206.20220208.001